A multiplexed antibody-based approach for analysis of denatured cellular proteins : development of western-MAP

(1)

0

(2)

1

ACKNOWLEDGEMENTS

The work for this thesis was carried out at the Department of Immunology, Rikshospitalet, Oslo University hospital from August 2012 to August 2013 supervised by researcher Fridtjof Lund- Johansen. Professor Tor Lea has been my supervisor at Norwegian University of Life Sciences.

During the work with this thesis I was supported with equipment and reagents from Expedeon and Abcam. This has been valuable for the execution of this thesis.

First, I would like to thank my brilliant supervisor, Fridtjof Lund-Johansen, who has been extremely helpful guiding me throughout the work of this thesis and always been available to give tips, guidance and support. Your great enthusiasm and dedication has been important.

I will also like to thank Professor Tor Lea for critical and valuable feedback during my writing period. I am very grateful for all help from all of you working at the Department of Immunology, especially Wei Wei Wu, Raquel Bartolomé, Marit Inngjerdingen and Grethe-Elisabeth Stenvik.

Without you I would be helpless considering practical matters in the lab. Thank you for the company, helpful conservations and the laughs during my stay.

Finally, I would like to give a big thanks to my family and friends for supporting and motivating me throughout this period of time.

Ås, August 15^th, 2013

Anette Lie Christensen

(7)

6

ABSTRACT

This thesis describes a new method for multiplexed analysis of cellular proteins. The cellular proteins were biotinylated and fractionated by gel-electrophoresis using a device (Gelfree 8100), which yields liquid fractions containing proteins of different sizes. Color-coded microspheres with antibodies to cellular proteins were added to the fractions. After incubation with the microspheres, captured proteins were labelled with fluorescent streptavidin and detected by flow cytometry. The results obtained with the assay resemble those obtained with standard western blotting. However, while western blotting is used to detect one or a few proteins at a time, the array-based assay developed here can be used to measure thousands of proteins simultaneously.

This approach is hereafter referred to as western-MAP (Microsphere-based Affinity Proteomics).

Variables that were found to affect performance of western-MAP included sample loading, choice of protein label, type of gels, addition of detergents and removal of free SDS from the fractions. The assay performance was optimal when the gels were loaded with 230μg of protein.

Combined labelling with amine- and thiol-reactive biotin was superior to either reagent used alone. Protein detection was enhanced when SDS in the protein fractions was removed by potassium chloride (KCl) precipitation, and further enhanced by the addition the non-ionic detergent Tween 20. The most useful gels contained 8% and 10% acrylamide.

The microsphere-based arrays that were used in this study contained thousands of antibodies.

Among these, 537 were found to capture a protein with a size compatible with that of the intended target. For 89 antibodies we had access to western blotting data obtained with the same cell lysates. A total of 76.4% of these antibodies showed comparable results between standard western blotting and the method developed in this thesis. Thus, the work resulted in successful multiplexing of western blot, which is one of the most widely used assays in protein research.

(8)

7

SAMMENDRAG

Denne avhandlingen beskriver en ny metode for multipleksanalyse av cellulære proteiner. De cellulære proteinene ble biotinylert og fraksjonert ved gel-elektroforese ved hjelp av et instrument (Gelfree 8100), som gir væskefraksjoner med proteiner av forskjellige størrelser.

Fargekodede mikropartikler med antistoffer mot cellulære proteiner ble tilsatt fraksjonene.

Proteiner bundet av antistoff ble merket med fluorescerende streptavidin og detektert ved flowcytometri. Resultatene fra denne metoden ligner de man får med tradisjonell western blotting, men til forskjell fra sistnevnte som kun kan analysere ett protein av gangen, kan metoden denne avhandlingen beskriver analysere flere tusen proteiner samtidig. Denne metoden blir heretter kalt western-MAP (Microsphere-based Affinity Proteomics).

Variabler som ble funnet å påvirke ytelsen til western-MAP var proteinkonsentrasjonen i prøven, valg av proteinmerking, type geler, tilsetning av detergent og fjerning av fritt SDS fra fraksjonene. Ytelsen var optimal når gelene ble lastet med 230μg protein. Kombinert merking av proteinene med både amin- og thiolreaktivt biotin var bedre enn å anvende en av disse reagensene alene. Fjerning av fritt SDS ved kaliumklorid (KCl) presipitering forbedret deteksjonen av proteinene. Når fraksjonene ble tilsatt den ikke-ioniske detergenten Tween 20, resulterte dette i lavere bakgrunn, og forbedret påvisning av membranproteiner. De mest nyttige gelene inneholdt 8 % og 10 % akrylamid.

De partikkel-baserte arrayene som ble brukt i denne avhandlingen inneholdt tusenvis av antistoffer. Blant disse, ble 537 funnet å binde et protein med en størrelse tilsvarende proteinet antistoffet var rettet mot. For 89 antistoffer hadde vi tilgang til western blott data fra de samme cellelysatene. Totalt 76,4 % av disse antistoffene viste lignende resultater i standard western blotting og metoden som ble utviklet i denne avhandlingen. Således, resulterte arbeidet i en multipleks versjon av western blot, som er en av de mest brukte metodene i proteinstudier.

(9)

8

1.0 INTRODUCTION

The overall aim of the work presented in this thesis was to develop an improved technology for large-scale analysis of cellular proteins. Studying proteins is fundamental for understanding cellular processes, as proteins are the functional components of all biological systems ¹.

Western blotting is one of the most widely used methods to study proteins in biochemical research. The technique is relatively simple to perform and has the advantage that it discriminates intended antibody targets from cross-reactive proteins. An important limitation is that the technique is limited to measuring one protein at a time.

During the past decades there has been a tremendous development in the field of large-scale protein analysis, or proteomics. This is largely due to advances in the field of mass spectrometry (MS). With modern MS it is possible to detect thousands of proteins in one sample. Studies based on the use of MS have greatly increased our understanding of cellular proteins. However, the protocols are complex and very time-consuming ^{2, 3}.

Antibody array analysis may provide a high throughput alternative to MS. Many attempts have been made to develop protocols for antibody array analysis ^4-9. In this assay format, the proteins in the sample are labelled with fluorescent dyes or haptens. Antibodies to proteins of interest are spotted onto predefined locations on a slide or bound to microspheres with fluorescent colour codes. The immobilized antibodies are used to capture labelled sample proteins. A limitation of the assay format is that the specificity is determined by the capture antibody alone. This is an important limitation since antibodies often bind more than one target. The issue is further complicated by the fact that many cellular proteins occur in a variety of multi-molecular complexes. Thus, even a mono-specific antibody can bind more than one protein. A review from 2002 describes antibody array analysis as a western blot where all the bands in the lane are compressed into one¹⁰.Thus, all cross-reactive binding contribute to the signal.

The goal for the work performed in this thesis was to develop an "antibody array western blot".

This could be achieved by fractionating biotinylated cellular proteins by gel-electrophoresis.

However, rather than blotting the proteins from the gel over to a membrane, they could be eluted into liquid fractions, using the Gelfree 8100 fractionation system from Expedeon Inc. The fractions will contain proteins with a narrow size distribution, and when a series of fractions is

(10)

9 analysed by antibody arrays, the results should be similar to those obtained by western blotting, except that thousands of antibodies could be used in parallel. We call this approach Western- MAP (microsphere-based affinity proteomics).

The primary goal for this thesis was therefore:

To develop a method for large-scale analysis for cellular proteins using microsphere-based affinity proteomics (MAP) combined with a method for size fractionation of denatured proteins.

To achieve this goal I first investigated if the Gelfree fractionation system is suitable for fractionating proteins prior to analysis with MAP. Next, I optimized the conditions to get the best results possible, before the new approach was compared to traditional western blotting.

(11)

10

2.0 BACKGROUND

2.1 HUMAN GENOMICS

The human genome consists of all the genomic information in humans. The first draft of the human genome (HUGO-project) identified the presence of between 30.000 and 40.000 protein coding genes ¹¹, but later studies have showed that the correct number is closer to 20.500 ¹². The genes provide a blueprint for all the building blocks that constitute the human organism: the proteins. Understanding the genes is therefore essential for understanding biology and disease.

There are a number of methods that allow genome-wide analysis. Microarrays have been available for more than two decades. The arrays consist of planar slides where fragments of DNA are spotted to predefined locations (Fig. 1). The DNA, cDNA or RNA in the sample to be analysed is labelled with fluorescence, and applied onto the slide. The fragments that are complementary to the fragments on the slide will hybridize, and the signal can be detected with a fluorescence scanner ¹³. DNA microarrays provided the first possibility to perform genome-wide analysis.

For the last approximately 35 years the Sanger method has been used for sequencing genomes ¹⁴. This is a resource- demanding method, and whole genome sequencing has therefore been limited to large sequencing centres. Almost ten years ago the first next-generation sequencing technology was commercialized. In the recent years these methods have evolved, and whole-genome sequencing can now be done in a matter of days. This has opened completely new possibilities for studying the human genome. The challenge today is not sequencing the genes, but analysing the tremendous amount of data acquired.

Fig. 1. DNA microarray. This is a schematic presentation of DNA microarray. A: Fragments of DNA is spotted onto predefined locations on a planar slide. B: The sample DNA or RNA are labeled with a fluorescent tag. C: The labeled sample DNA or RNA is applied to the slide, and fragments that are complimentary to fragments on the slide will hybridize. D: The fluorescence signal from the hybridized sample fragments can be detected.

(12)

11

2.2 HUMAN PROTEOMICS

Genomics provide information about the primary structure of proteins. However, proteins are more complex. Eukaryotic genes have both protein coding exons and non-coding introns, and the introns must be removed from the mRNA transcripts before translation. This process is called RNA-splicing, and a given mRNA can often be spliced in different ways to provide multiple products. This is called alternative splicing and results in that a given gene can give rise to more than one protein species. About 95% of multiexon human genes are subject to alternative spicing¹⁵.

After translation the proteins can undergo post translational modifications (PTM). Over 300 such modification exists, and new modifications are discovered regularly ¹⁶. The most important PTMs are phosphorylation, glycosylation, formation of disulphide bridges or cleaving of a pro- peptide.

Due to alternative splicing, there are more protein species than protein coding genes, and PTMs give rise to an even greater diversity (Fig. 2). It has been suggested that the human genome potentially can produce 1.8 million different protein species ¹⁶. Today the Uniprot database has more than 23.000 reviewed entries for human proteins ¹⁷.

The genome is relatively stable, whereas the transcriptome and the proteome are constantly changing. The transcriptome is all the mRNAs in an organism, a cell or a population of cells at a

Fig. 2. The complexity increases from the genome to the proteome. Many transcripts (mRNAs) can be produced from one single gene due to alternative splicing, and not all mRNAs are translated into proteins. Post translational modifications can result in several different protein species from the same mRNA. As a consequence of this, humans have a greater number of proteins than the number of protein coding genes. The exact number of proteins that make up the human proteome remains unknown, but up to 1,8million protein species has been suggested ⁽¹⁵⁾.

(13)

12 given time, under given circumstances, and the proteome is the equivalent for proteins. The transcriptome can to a certain extent be used to predict protein levels, however not all mRNA is translated, and different mRNA is translated and degraded at differential rates. It is possible to envision the genome as what the cell knows how to do, the transcriptome as what the cell is thinking about doing and the proteome as what it is currently doing. It is fair to say that while genes encode biological systems, the proteins are biological systems. It is therefore essential to study proteins to understand the cellular systems and processes.

Initiatives similar to the human genome project have now been taken to characterize the human proteome. This project is called the human proteome project (HPP). This project may seem like a logical extension of the HUGO-project, but it has been questioned if the current technology is ready to handle such an immense task ¹⁸. The experimental strategy suggested for the human proteome project is to use three working pillars: MS, antibody capture and immunohistochemistry and bioinformatics ¹⁹. Even though there has been a tremendous development in all these fields, the technology is not nearly as robust as those used to sequence the genome.

2.3 PROTEIN SEPARATION

The proteome is highly complex, and identifying individual proteins in complex biological samples is therefore a difficult task. To simplify the task, the samples are often fractionated to reduce sample complexity. Fractionation methods are therefore fundamental for protein analysis.

The proteins molecular mass is a highly predictable parameter for fractionation, since it is usually unaffected by sample or solvent conditions.

2.3.1 ELECTROPHORESIS

Polyacrylamide gel electrophoresis (PAGE) is a widely used approach for fractionating proteins according to their molecular weight (MW). Electrophoretic fractionation is based on the principle that charged molecules migrate under the influence of an electric field. PAGE is carried out by adding the samples to wells on the polyacrylamide gel and applying an electrical current over the gel. The migration of the molecules in the gel is influenced by the size, shape and net charge of the molecules ^{20, 21}. The difference in migration results in molecules with similar properties ends up at the same place in the gel, called a band.

(14)

13 Gel electrophoresis can be carried out with both native and denatured proteins. When performing gel electrophoresis with native proteins, it is not possible to distinguish between the effects of the size, shape and net charge to the migration speed through the gel. This means that proteins with different MWs can have the same mobility in the gel, and end up in the same band.

This limitation with native-PAGE is overcome in SDS-PAGE. Here the proteins are heated with the anionic detergent sodium dodecyl sulphate (SDS) prior to electrophoresis. SDS binds most proteins, and provides them with a negative charge. Binding of SDS also causes denaturation of proteins by disrupting all the non-covalent protein bonds. The results is that all proteins migrate towards the anode, and the migration speed will only depend on the size of the protein ²⁰.

SDS disrupts all non-covalent protein bonds, but it is also desirable to break the disulphide bonds.

Disulphide bonds are covalent connections that can form between thiol-groups on two cysteine- residues in a protein, or between two cysteine-residues on separate proteins. Disulphide bonds are broken by treating them with reduction agents. The most widely used reduction agents are DTT (dithiothreitol) and β-mercaptoethanol, but in the last decade TCEP (tris(2- carboxyethyl)phosphine) has been available ²². An advantage of TCEP is that the agent does not contain thiols and is therefore compatible with thiol-reactive protein labels ²².

2.3.2 TWO DIMENSIONAL GEL ELECTROPHORESIS (2DE)

To fractionate the samples into even simpler factions two dimensional electrophoresis (2DE)²³ is used. In 2DE the proteins are first fractionated by isoelectric focusing (IEF) in the first dimension. Typically IEF occurs in a gel strip containing a pH gradient. The proteins will stop migrating in the gel when it reaches the pH value that corresponds to their isoelectric point (pI).

pI is the pH where the net charge of the protein is zero. The charge of the protein depends on the residuals in the protein, as well as the pH of the environment. After the first dimension the IEF strip is layered over an SDS-PAGE to fractionate proteins by MW (Fig. 3.). The proteins in the gel may be detected by silver staining or fluorescent labelling ²⁴.

(15)

14 While 2DE is highly useful to

fractionate proteins, identification relies on other methods. The most commonly used detection method for proteins fractionated with 2DE today is MS. To analyse the proteins in the gel with MS or other detection methods, the protein spots have to be cut out of the gel and treated with a proteolytic enzyme, usually trypsin, to form peptides that elute from the gel ²⁴.

Two dimensional electrophoresis followed by in-gel protein digestion and elution is a slow and laborious method. Problems are also associated with elution and solubility, especially with hydrophobic proteins ²⁵, even though detergents ^{26, 27} have increased the representation of membrane proteins in the results. Moreover, the resolution provided by the electrophoretic fractionation step is limited. Proteins with very high or very low MW and proteins with extreme pIs are usually not detected using standard 2DE, but improvements have been obtained using wide pH gradients ²⁸ and different buffer systems ²⁹. The last limitation with 2DE is the ability to detect low abundance proteins. The dynamic range of 2DE is about 10⁴, but the protein expression in human cells is estimated to be 7-8 orders of magnitude.

2.3.3 CHROMATOGRAPHY

Chromatography is a widely used approach for fractionation of native proteins. In chromatography, proteins in solution are fractionated by their migration pattern across an immobile matrix. The solution is referred to as the mobile phase and the immobile matrix as the stationary phase. The basis for the fractionation is that molecules migrate at differential rates depending on the affinity for each of the two phases.

There are many types of chromatography, and the various approaches use different properties of the peptides or proteins for fractionation. Hydrophobic interaction chromatography (HIC)

Fig. 3. Two dimensional electrophoresis (2DE). This is a schematic presentation of 2DE. In the first dimension the proteins are fractionated using a strip with a pH gradient, witch fractionates the proteins according to pI. This is called isoelectric focusing (IEF). After IEF the strip is layered over an SDS-PAGE, and the proteins are fractionated according to molecular weight in the second dimension.

(16)

15 fractionates peptides or proteins according to their polarity, while in ion exchange chromatography (IEC) the proteins are fractionated and purified on the basis of ionic interactions.

Affinity chromatography fractionates proteins on the basis of reversible biological interactions, such as antibody-antigen interactions.

Size exclusion chromatography (SEC) is used to fractionate proteins according to molecular weight. The resin in the column consists of porous particles. Large molecules cannot enter into the pores and elute first, while small molecules pass in and out of the pores through the column, and elute later. While SDS-PAGE is used to separate denatured proteins, SEC is typically performed under native conditions. A given protein species may therefore elute in non- overlapping fractions depending on whether it occurs as a monomer or in one or more complexes.

SEC is therefore often used to analyse protein complexes.

2.3.4 CONTINUOUS ELUTION TUBE GEL ELECTROPHORESIS In continuous elution tube gel

electrophoresis, proteins are separated by gel-electrophoresis, but are finally eluted into liquid fractions. Several such approaches have been developed and used in research (for example ^30-34). The utility of this method has been limited by the fact that it has been biased towards the lower MW range. Moreover, the fractions are subject to large dilution, especially with the fractions containing proteins with high MW, and the fractionation is time consuming.

To overcome the limitations of traditional techniques, Tran and Doucette developed a device that uses a short gel column for protein fractionation where the proteins

Fig. 4. Gelfree cartridge. This is a schematic presentation of a Gelfree cartridge. The gel columns, the cathode and anode buffer reservoir, the sample loading and the sample collection chamber make up the five main parts. The electrodes are inserted into the cathode and the anode buffer reservoir, the sample is loaded into the sampled loading chamber and electrical current is applied. The proteins will now migrate through the gel column and end up in the sample collection chamber, where fractions can be collected at predefined intervals. The smallest proteins will migrate fastest through the gel, and end up in the first fractions, while the larger proteins will end up in the later fractions.

(17)

16 ultimately are eluted from the column, and can be collected in solution. The separation technique was named Gelfree for Gel-Eluted Liquid FRaction Entrapment Electrophoresis³⁵. The Gelfree fractionation cartridges consist of five main parts: a cathode and an anode buffer reservoir, a sample loading and a sample collection chamber and the gel column (Fig. 4). The anode, cathode and sample collection chamber are filled with running buffer and the sample are loaded onto the sample loading chamber. When electrical current is applied the proteins will migrate through the gel column, and are fractioned according to MW, before they finally elute into the collection chamber. The fractions are then collected from the collection chamber at defined time intervals ^35-

37. The short column makes the separation more rapid ³⁷, and the protein recovery is high ³⁵. Botelho et al. performed a MS experiment to compare the Gelfree fractionation system to a more traditional approach involving proteolytic digestion of proteins in bands cut out from SDS- PAGE gels, and concluded that the two methods yield comparable results in both type and number of proteins identified ³⁸. The Gelfree system has also been used to prepare samples for MS with good results in a number of other studies ^38-46.

2.4 ANTIBODY-BASED METHODS

Antibodies are widely used to detect proteins. These reagents provide means to detect proteins in complex samples with high throughput and precision. As explained below, however, antibody- based methods are largely limited to detecting one or a few proteins simultaneously.

2.4.1 ANTIBODIES

Antibodies are glycoproteins produced by B-cells and are part of our immune system. B-cells carry antibodies on their surface, while activated B-cells, called plasma cells secrete soluble antibodies. Antibodies bind antigens specifically. Antigens are molecules which stimulate B-cells to antibody production, and the binding site is referred to as an epitope. A linear epitope is a result of a contiguous amino acid sequence (Fig. 5), whereas a conformational epitope consists of non-neighboring amino acids brought in proximity by protein folding. Conformational epitopes will be destroyed by denaturation, while some linear epitopes can be masked when the protein is in its native form and can therefore only be detected when the protein are denatured.

(18)

17 Antibodies consist of four polypeptide

chains that are covalently linked together with disulphide bonds (Fig. 6). The four polypeptide chains are divided into two heavy and two light chains. There are two types of light chains called κ and λ, and five types of heavy chains called α, μ, γ, δ and ε.

The five different types of heavy chains, give rise to five classes of antibodies: IgA, IgM, IgG, IgD and IgE. These five different

classes have different structural and functional properties, and IgG is the class most often used in antibody-based methods, since it has the highest affinity for their antigens.

Antibodies used in antibody-based methods can be divided into two major groups: polyclonal and monoclonal antibodies (Fig. 6). Polyclonal antibodies are obtained from serum of immunized animals, and may recognize different epitopes on the same antigen. Monoclonal antibodies have specificity for only one epitope, and are made by hybrid cells, which are generated by a fusion of B-cells and myeloma cells. The fusion gives rise to cells that have the antibody specificity of the B-cells, and self-renewing capacity of the myeloma cell. These cells can be grown in large quantities, and the technology is called hybridoma technology. In 1984 Milstein and Kohler received the Nobel Prize for the discovery and production of monoclonal antibodies ^47-49.

Production of antibodies by immunization of animals is time-consuming and expensive. Hybridoma technology has for the past 35 years enhanced research and diagnostics by providing monoclonal antibody reagents. In 1990 McCafferty described a technique for in vitro production of monoclonal antibodies ⁵⁰. In this technique the variable immunoglobulin genes from B-cells are

Fig. 5. Conformational and linear epitopes. This is an illustration of an antigen. Epitopes are the precise parts of an antigen where the antibody binds. Epitopes can be

conformational or linear. Conformational epitopes is made as a result of the folding of the amino acid chain, while linear epitopes consist of a contiguous sequence of amino acids.

Fig. 6. Antibody. An antibody consists of four polypeptide chains, two heavy and two light chains, which are linked together with disulphide bonds.

(19)

18 amplified by polymerase chain reaction (PCR), and expressed

on the surface of bacteriophages. Phage particles with inserts encoding specific binding domains are selected by using a chromatography column with immobilized antigen. This technic permits control over selection and screening conditions, and therefore allows production of antibodies against defined epitopes. With in vitro display techniques it is possible to target molecular structures that are not easily recognized during an immune response in vivo ^51-56.

2.4.2 WESTERN BLOTTING

One of the most widely used antibody-based methods today is western blotting. This technique was developed about 30 years ago ⁵⁷, and is applicable for a wide range of samples including serum, tissue culture supernatants, cell and tissue extracts.

The first step of western blotting is SDS-PAGE or native-PAGE where the proteins are fractionated by MW. Then the proteins in the PAGE-gel are transferred, or blotted, onto a membrane. During transfer the gel and the membrane are placed on top of each other in a buffer, and an electrical current is applied to cause the proteins to migrate. After transfer, the membrane is treated with agents that block non-specific binding sites, before it is labelled with antibodies. In a typical experiment, binding of the primary antibody to the proteins on the membrane is detected using a secondary antibody, which is conjugated to an enzyme such as horse radish peroxidase. A substrate is added, and the enzyme generates a product that gives a signal, usually chemiluminescence (Fig. 8).

Antibody reactivity on western blots is observed as bands, and in many cases there are many antibody stained bands on the membrane. The position of the band corresponding to the intended target, however, is predictable from the size of the protein. Moreover, the specific binding is usually more consistent among different samples than the cross-reactivity. The relative intensity of bands from different samples provides semi-quantitative information about the amount of the protein. The limitation with western blotting is that the technique can only be used to detect one

Figure 7. Polyclonal and monoclonal antibodies. This is an illustration of the difference between polyclonal and monoclonal antibodies. A: Polyclonal antibodies bind different epitopes on the same antigen. B: Monoclonal antibodies have specificity for only one type of epitope.

(20)

19 or a few proteins at a time.

2.4.3 FLOW CYTOMETRY

Flow cytometry is used to analyse cells in suspension.

In diagnostics flow cytometry is commonly used to count and characterize subsets of leucocytes. This is carried out with supravital stained cells that are introduced to the flow cytometer by a fluidics that focus the cells into a capillary. The capillary is intersected by one or more lasers, and detection of scattered laser light is used to estimate cell shape, cell size, cellular granularity, nuclear lobularity and cell surface structure.

The cells can also be labelled with fluorochrome- conjugated antibodies. Multiple antibodies with different fluorescent probes can be used at the same time ⁵⁸. The lasers excite the fluorochromes on the antibodies, and the emitted light is collected through lenses and guided via fibre optics to detectors.

Fluorescent dyes that are excited by the same laser and emits light at different wavelengths are discriminated by the use of optical filters. Dyes with similar emission, but different excitation can also be discriminated since they pass the lasers at different time points, and the instrument can resolve the time difference.

Fig. 8. Western blotting. This is a schematic presentation of the work flow in western blotting.

First the proteins are applied to wells on a PAGE gel and electrical current is applied. The small proteins will migrate faster than bigger proteins, resulting in fractionation of the proteins according to MW. After PAGE the proteins are transferred to a membrane, and antibodies are added. The primary antibody binds the target and a secondary antibody conjugated with an enzyme targets the primary antibody. A substrate, which is converted to a chemiluminescent compound by the enzyme, is added, and the signal can be detected.

(21)

20 2.4.4 ANTIBODY SANDWICH ASSAYS

All the methods described so far in this thesis are used to detect immobilized proteins. Antibody sandwich assays detect proteins in solution. The term ELISA (Enzyme-linked immunosorbent assay) is used to describe all forms of enzyme immunoassays with colorimetric detection principle. When performing ELISA, capture antibodies are immobilized on a solid matrix, typically a microtiter plate, the samples are added and the proteins from the sample bind to the capture antibodies. A second antibody, binding a different epitope on the same protein, is used for detection. Typically this antibody is conjugated with an enzyme such as peroxidase, and the signal is detected using a substrate that is converted to a coloured or chemiluminescent compound. The ELISA format provides dual specificity since the capture and detection antibodies target two distinct epitopes. This makes this method less affected by cross-reactivity.

There are a large number of different variants of antibody assays with different types of solid matrix, different types of antibody arrangement and different types of detections methods.

Different immunological assays are widely used both in diagnostics and in research. The solid matrix can be covered with antibodies or antigens, and can therefore bind antigens or antibodies in solution. The solid matrix can be different types of beads or slides instead of a microtiter plate.

The beads can have diverse features such as being paramagnetic or have the ability to be coloured with fluorescents labels, giving the advantage that the beads can be coloured differently, and several antibodies can be used at the same time. Paramagnetic beads are often used in automated diagnostic methods to bypass the need for centrifugation.

2.4.5 LIMITATIONS WITH ANTIBODY-BASED METHODS

Antibody-based methods are inexpensive, rapid and simple to preform, but some important limitations apply. The assays are largely limited to detecting a small number of proteins at a time.

Equally important is the fact that well characterized reagents mainly cover a relatively low number of proteins that have been studied extensively over many years.

(22)

21 An initiative for producing antibodies towards all human

proteins has been taken ⁵⁹. This is called the HUPO antibody initiative. It has been suggested that an objective for antibody- based proteomics should be to generate renewable, preferably monoclonal, paired antibodies towards all human proteins ⁵⁹. Paired antibodies are two or more antibodies recognizing distinct and non-overlapping epitopes on the same antigen (Fig.

9).

Antibodies may bind other proteins than the one used for immunization. This is commonly referred to as cross-reactivity

or off-target binding. Antibodies can react with similar, related epitopes on proteins from the same gene family, but also to completely unrelated epitopes ^{60, 61}. This gives results that are difficult to interpret. Cross-reactivity is a big problem when using polyclonal antibodies, but can potentially be a problem when using monoclonal antibodies too. A way of limiting cross- reactivity is to use paired antibodies.

Exactly how frequent cross-reactivity occurs is difficult to say, but in a study performed by Schwenk et.al.⁶² only 531 out of 11,000 antibodies showed a single band in western blotting. In another study the proteins captured by anti-TSP1 was analysed by MS ⁶³. The study showed that TSP1 (thrombospondin-1) was the major target, but other proteins were also detected.

Antibody performance is application-dependent. To some extent this is because the proteins are denatured by detergents or formalin in many applications, meaning that conformational epitopes are destroyed. In these types of applications antibodies directed against linear epitopes should be used. For applications where the proteins are in their native form, the epitopes can be masked due to the folding of the protein, or due to protein-protein interactions. A study of more than 5000 antibodies from over 50 commercial providers showed that as many as 50% of the antibodies where non-functional in the immunohistochemistry application used ⁶⁴.

In all the methods except for ELISA, there are no standardized measurements for quality control.

Researchers must therefore frequently perform their own evaluation ⁶⁵. As a part of the HUPO antibody initiative a web portal called Antibodypedia has been developed ^{59, 64}. Antibodypedia is

Fig. 9. Paired antibodies. Cross- reactivity can be minimized when using paired antibodies in antibody arrays. Paired antibodies mean two or more antibodies that recognize distinct and non-overlapping epitopes on the same antigen.

(23)

22 a portal where scientists can share information about validation of antibodies. Today Antibodypedia contains 810,580 reviewed antibodies from 43 providers, covering gene products from 18,831 protein coding genes, which is about 91% of all human genes ⁶⁶. The aim of Antibodypedia is to make a resource for validated antibodies towards all human proteins ⁶⁴.

2.5 MASS SPECTROMETRY

Mass spectrometry provides means to measure the mass and charge of peptides and peptide fragments. Mass spectrometers can examine peptides of up to 50 amino acids in length, but since most proteins are longer than this, the proteins are treated with a proteolytic enzyme prior to analysis. The standard approach is to digest the proteins with a sequence specific protease, such as trypsin. Trypsin cleaves the proteins immediately after arginine or lysine residues, and since the amino acid sequences of all human proteins are generally known from HUGO-project, digestion of any protein will result in a predictable set of peptides with known mass. The peptide masses obtained from a MS experiment can therefore be used to search different bioinformatics databases to identify the protein.

Mass spectrometric measurements are carried out on ionized analytes in the gas phase. Protein and peptides are large non-volatile molecules, but there are two approaches to make gas phase ions from proteins: matrix-assisted laser desorption/ionization coupled to time-of-flight analysers (MALDI-TOF)⁶⁷and electrospray ionization (ESI) ⁶⁸. When using MALDI-TOF the proteins are mixed with a matrix, which contains small organic solvents and is responsible for ionization of the proteins. The matrix is dried out, and energy from a laser makes the peptides in the matrix go into gas phase. Then the mass-charge ratio is measured. This is done by measuring the time a peptide uses from the place of ionization to the detector (Time-of-flight, TOF).

When using ESI for ionization a strong electrical field is applied, under atmospheric pressure, to a liquid passing through a capillary tube. This leads to highly charged droplets, which in turn causes the ions to be separated from the solvent. When using ESI the samples start out as liquid, so this approach is often combined with different types of liquid chromatography (LC-MS).

There are a number of different mass analysers. Along with TOF, which is already mentioned, the ion trap, quadrupole and Fourier transform ion cyclotron are currently used in research.

(24)

23 A tandem mass spectrometer or MS/MS consists of two mass analysers separated by a collision cell. The first mass analyser is set to only allow a single type of peptide with a specific mass to charge ratio to continue into the collision cell. In the collision cell, the peptide is fragmented even further. This fragmentation occurs primarily at peptide bonds, and gives therefore ions that only differ by a single amino acid in mass. Since almost all amino acids have different mass, the recorded masses can be used to determine the amino acid sequence of the protein.

MS-based proteomics has been very important for much of our knowledge regarding proteins and protein activity. Among others the approach has been used to determine the proteome of organelles, the protein composition of a cell, members of protein complexes and post translational modifications ^{69, 70}.

In “shot-gun” or “discovery” proteomics, the aim is to identify all the proteins in a sample. This is challenging, since the number of peptides that are formed when complex samples, such as serum or cell lysates, are digested with enzymes by far exceeds the number that the mass spectrometer can resolve. Extensive fractionation is therefore necessary prior to analysis.

MS is still a technically difficult method which requires a high level of expertise by the users.

This is well illustrated by a study where 27 laboratories received a test sample containing a mixture of 20 purified proteins. Only 7 laboratories were able to identify all proteins in this simple mixture ³. In another study more than 1000 proteins were detected reproducibly in serum samples from patients with cardiovascular disease ⁷¹. This was achieved after depletion of 12 abundant proteins and extensive fractionation of the remaining protein mixture, using 2800 hours of instrument time. While it is impressive to detect such a large number of proteins by MS, the actual number of proteins is likely to be at least ten-fold higher. No proteome has yet been completely analysed, and it will be difficult to determine when that milestone has been achieved, since a suitable reference does not exist ⁷².

(25)

24

2.6 ANTIBODY ARRAY ANALYSIS

To overcome limitations with antibody-based assays, attempts have been made to develop platforms similar to those used in genomics 4-8, 10, 73-76

. Collectively, these can be referred to as antibody array analysis.

Antibody arrays can be broadly divided into two categories. The first category is an assay format where immobilized antibodies are used to capture labelled sample proteins (Fig. 10A) ^{4, 77}. Another approach is to use dual specificity sandwich arrays, where the proteins first are captured by one antibody and then detected using another antibody against a different epitope on the same target (Fig. 10B) ^78-80. It is possible to label the detection antibodies directly or using two different antibodies for detection: one detection antibody that binds the target and one read-out antibody that binds the detection antibody.

Direct labelling single capture antibody assays are simple to perform, and the number of proteins that can be analysed is nearly unlimited, but there are problems associated with obtaining specificity ^{4, 10} Assays where immobilized

Fig. 10. Different experimental formats for antibody microarrays. This is a schematic presentation of two different antibody assay formats. A: In direct labelling, single-capture antibody array experiments the proteins in the sample is labelled prior to analysis. The capture antibodies are immobilized on the solid matrix and the samples are applied. The capture antibodies bind the labelled proteins and the signal is detected. B: This shows dual-antibody sandwich microarrays. The proteins from the sample are captured by the captures antibodies which is immobilised on the solid matrix. Detection antibodies, which provide dual specificity to the target protein, are added. Binding of the detection antibody to the target is detected using a labelled read-out antibody.

(26)

25 antibodies are used to capture labelled sample proteins are reliable when each capture reagent binds a single target only. However, mono-specific antibodies are exceedingly rare. Dual- antibody sandwich immunoassays have a much higher specificity, as the target is recognized by two different antibodies. On the other hand it can be difficult to find paired antibodies for the microarray. This type of array is also limited to analyse about 50 targets simultaneously⁷⁴, because the possibility for cross reactivity between the different detection antibodies increases with higher numbers of analytes ⁸¹.

Different types of solid matrixes can be used in microarrays, but planar glass or silicon arrays or microbead arrays are most common. In planar glass ^{4, 79, 80} or silicon ⁶ arrays the capture reagents are spotted onto slides at predefined locations, as in DNA microarray. This approach requires access to sophisticated printing devices. An alternative is to use microspheres with fluorescent colour codes ^{7, 9}. These are typically analysed by flow cytometry, and the assays can be

extensively multiplexed by using microspheres with different fluorescent labelling and different sizes.

The limitations associated with antibody microarrays compared to MS are that the assays depend on the availability of affinity probes. In antibody microarray it is also necessary to decide which proteins that are going to be studied prior to the experiments, MS allows unbiased analysis.

2.6.1 PROTEIN LABELLING

While sample proteins can be labelled directly with reactive forms of fluorescent dyes, earlier studies have shown that indirect labelling with biotin and streptavidin is superior ⁸². Biotin is a small molecule that binds to streptavidin with high affinity. Streptavidin can in turn be conjugated to detectable molecules such as fluorescent dyes.

The two most commonly used forms of biotin react with amine- and thiol groups in proteins. N- Hydroxysuccinimide (NHS) esters react with primary amino groups (-NH2). Primary amines are found in the side chain of lysine residues and the N-terminus of the polypeptide.

Maleimide derivatives react with thiols, found in cysteine residues. Cysteines are unstable in non- reducing environments where they form di-sulphide bonds with each other. For extracellular proteins, which occur in non-reducing environments, this is an important mechanism for protein

(27)

26 folding. When cysteines are reduced and reacted with maleimides, they can no longer form disulphide bonds. Thus, this type of labelling prevents refolding of denatured proteins. This is advantageous for the approach used in this thesis, where the aim is to detect and analyse denatured proteins. On the other hand, most proteins contain fewer cysteines than lysines, and approximately 4% of proteins lack cysteines.

2.6.2 WESTERN-MAP

In this thesis the aim was to develop an approach for multiplexed large scale analysis of cellular proteins where the results should resemble the results obtained with western blotting. The development of this approach was based on a previously published approach called SEC-MAP (Size-Exclusion Chromatography – Microsphere-based Affinity Proteomics) ^{9, 83-85}. In SEC-MAP the samples are fractionated with SEC prior to detection and analysis with antibody microsphere arrays. Microspheres give the opportunity to use microtiter plates with samples, and this gives a higher sample throughput compared to microarray slides.

The experimental format for the antibody microsphere array used in SEC-MAP is direct labelling single capture antibody. As already discussed direct labelling single capture antibody arrays have potential to detect and analyse far more proteins than the other antibody microarray approaches mentioned.

When using direct labelling single capture antibody arrays it is necessary to have a way of controlling the specificity. In western blotting the samples are fractionated according to MW prior to detection and analysis with antibodies. This separation makes it possible to distinguish unspecific bands from the intended targets. This approach can be adapted to microarrays, and is precisely what is done when preforming SEC-MAP. The sample fractionation, along with several antibodies against the same target in the same array, makes it possible to identify specific binding of proteins, and rule out cross-reactive binding ⁸⁵.

In traditional western blotting the proteins are denaturated with heat and reduction agents prior to fractionation. This gives a different fractionation pattern to what is obtained with SEC, witch fractionate native proteins. To develop an approach that would resemble western blotting it was therefore necessary to investigate if SEC could be replaced by another fractionation approach that fractionated denaturated proteins.

(28)

27 Fractionation with denaturated proteins could give several benefits. First of all protein complexes are disrupted during denaturation, meaning that all proteins will mainly occur an in monomeric form. The proteins´ MW are also the only factor affecting the protein fractionation, as opposed to fractionation with native proteins where shape and surface charge also can have an influence.

This makes the results easier to interpret. Many antibodies are developed for use in western blotting, as this is a widely used application. A study showed that antibodies developed for western blotting performed better when the samples were treated with heat ⁶². A limitation with denaturation is that the anionic detergent SDS is commonly used for this purpose, an in microarrays SDS has been shown to affect the results obtained ^{83, 86}. Antibodies directed at

Fig. 11. Western-MAP work flow. The approach for large scale analysis of cellular proteins developed in the thesis is called Western-MAP. To preformed western-map first lysates is made from the sample cells. The proteins in the sample are the labelled with biotin, and then fractionated. The three fractionation systems examined in this thesis are the Gelfree 8100 fractionation system from Expedeon (shown), the whole gel eluter from Biorad and Size-exclusion chromatography with denaturated proteins. After fractionation the antibody arrays is added to each fraction. The antibody array consists of antibody covered beads witch are coloured with fluorescent dyes. Each type of antibody had its microspheres coloured with a unique colour-code and/or size, and thereby the fluorescent colouring of the microspheres together with the microsphere size works as a barcode for the type of antibody bound. The antibody in the arrays then binds their target proteins from the fractionated sample. The antibody array microspheres are then collected, and the captured protein is labelled with streptavidin-PE. A flow cytometer is used to detect the PE-signal of any captured protein and the size and colour-code of the microspheres to identify the protein bound.

(29)

28 conformational epitopes will also be non-functional when the proteins are denatured

If SEC could be replaced with a fractionation approach that fractionated denaturated proteins, this would theoretically be a multiplexed approach that gave the same results as western blot (Fig.

11). This approach will be called western-MAP.

(30)

29

3.0 MATERIALS

3.1 CELL LINES

HeLa (Human epithelial carcinoma cell line) (Abcam, Cambridge, United Kingdom) Jurkat (Human T-cell lymphoblast-like cell line) (Abcam, Cambridge, United Kingdom)

3.2 REAGENTS AND CHEMICALS

Acetic Acid (Acetic acid (glacial) 100%, cat.no:

1000632500, Merck Millipore, Darmstadt, Germany)

Albumin (Albumin 40 mg/ml, cat.no: 478198,

octapharma, Lachen, Switzerland)

β-mercaptoethanol (2-Mercaptoethanol, cat. no: M7522 Sigma, Sigma-Aldrich, St. Louis, Missouri, USA) Coomassie Brilliant Blue R-250 (cat.no: 161-0400, Bio Rad, Hercules,

California, USA)

DMSO (Dimethyl sulfoxide, cat.no: D4540 SIGMA,

Sigma-Aldrich, St. Louis, Missouri, USA)

DTT (Dithiothreitol (DTT), cat.no: D-1532,

Eugene, Oregon, USA)

EDTA (Titriplex III, cat.no: 108418, Merck

Millipore, Darmstadt, Germany)

Foetal bovine serum (cat.no: F6178 SIGMA, Sigma-Aldrich, St.

Louis, Missouri, USA)

Glycerol (Glycerol solution, cat.no: 49782, Sigma-

Aldrich, St. Louis, Missouri, USA)

(31)

30

HEPES (cat.no: H3375 SIGMA, Sigma-Aldrich, St.

KCl (cat.no:104936, Merck Millipore, Darmstadt,

Germany)

Lauryl maltoside (LM) (n-Dodecyl β-D-maltoside, cat.no: D4641 Sigma, St. Louis, Missouri, USA)

Maleimide-biotin (EZ-Link Maleimide-PEG2-Biotin, cat.no:

21901 Thermo Fisher Scientific, Rockford, Illinois, USA)

Methanol (cat.no: 32213 Sigma-Aldrich, Sigma-

Aldrich, St. Louis, Missouri, USA)

MgCl2 (cat.no: 63065 Fluka, Sigma-Aldrich, St.

NaCl (cat.no: 106404, Merck Millipore, Darmstadt,

Germany)

NaF (cat.no:S7920 Sigma-Aldrich, Sigma-Aldrich,

St. Louis, Missouri, USA)

Na₂HPO₄*12H₂O (cat.no: 106579, Merck Millipore,

Darmstadt, Germany)

NaH₂PO₄*H₂O (cat.no: 106346, Merck Millipore, Darmstadt, Germany)

NA3VO4 (Sodium orthovanadate, cat.no: S6508 Sigma,

Sigma-Aldrich, St. Louis, Missouri, USA)

(32)

31

NHS-biotin (EZ-Link NHS-PEG₄-Biotin, cat.no:21363,

Thermo Fisher Scientific, Rockford, Illinois, USA)

PMSF (Phenylmethanesulfonyl fluoride, cat.no:

P7626 Sigma, Sigma-Aldrich, St. Louis, Missouri, USA)

SDS (Sodium dodecyl sulphate, cat.no:

L3771 SIGMA, Sigma-Aldrich, St. Louis, Missouri, USA)

Sephadex G25 Fine (cat.no: 17-0032-01, GE Healthcare, Uppsala, Sweden)

Sigma protease inhibitor cocktail (cat.no: P8340 Sigma, Sigma-Aldrich, St.

Streptavidin-Phycoerythrin (R-phycoerythrin conjugated with streptavidin, cat. no: 016-110-084, Jackson Immunoresearch, West Grove, Pennsylvania, USA)

TCEP (Tris(2-carboxyethyl)phosphine

hydrochloride, cat. no: C4706 Aldrich, Sigma-Aldrich, St. Louis, Missouri, USA)

Tween 20 (cat. no: P1379, Sigma-Aldrich, St. Louis,

Missouri, USA)

3.3 COMMERCIAL SOLUTIONS

Casein blocker in PBS (cat.no: 37582, Thermo Fisher Scientific, Rockford, Illinois, USA)

Gelfree 8100 HEPES Running Buffer (cat.no: 42202, Expedeon, Cambridgeshire,

(33)

32 United Kingdom)

Gelfree 8100 Tris Acetate Sample Buffer (cat.no: 42302, Expedeon, Cambridgeshire, United Kingdom)

Precision Plus Protein Dual Colour Standard (cat.no: 161-0374, Bio Rad, Hercules, California, USA)

RPMI-medium (RPMI 1640 Medium, GlutaMAX, cat.no:

61870-010, Life Technologies, Paisley, UK)

RunBlue LDS Sample Buffer (cat.no: NXB31010, Expedeon,

Cambridgeshire, United Kingdom)

RunBlue SDS Run Buffer (cat.no: NXB50500, Expedeon,

Cambridgeshire, United Kingdom)

3.4 SOLUTIONS PREPARED IN THE LAB

Coomassie blue staining solution 0.5g Coomassie blue R-250 800mL Methanol

140mL Acetic Acid

Deionized water (dH₂O)up to 2.0L

The components were mixed together, and the solution was stored at room temperature (RT).

Destaining solution 300mL Methanol 100mL Acetic acid dH2O up to 1.0L

The components were mixed together, and the solution was stored at RT.

(34)

33 Lysis buffer

20 mM HEPES, pH8.0 2 mM MgCl₂

1 mM EDTA 1mM NaF 1mM Na3VO4 140mM NaCl

1% Lauryl Maltoside 1mM TCEP

1mM PMSF

1mM Sigma protease inhibitor cocktail

The components were dissolved in dH2O, and stored at 4°C.

Maleimide-biotin (10mg/mL)

10mg dissolved in 1mL DMSO and stored at -20°C.

NHS-biotin (60mg/mL)

60mg/mL NHS-biotin was dissolved in 1mL DMSO Phosphate buffered saltwater (PBS) x25

4.48g NaH₂PO₄*H₂O 48.45g Na₂HPO₄*12H₂O 204.5g NaCl

The components were dissolved in 1liter of dH₂O, and stored at RT. PBS was made from the 25xPBS stock by taking 40mL stock and diluting it to 1liter in dH2O.

PBS-1%tween 40mL PBS 10mL Tween 20

The components were dissolved in 1L of dH2O, and stored at RT.

Sephadex G25 solution

7g of Sephadex G25 was added 50mL dH2O, and stored at RT.

(35)

34 Tween 20 (10%)

1mL Tween 20 was dissolved in 9mL dH₂O, and stored at 4°C.

3.5 ANTIBODIES

Goat gamma globulins (cat. no: 005-000-003 Jackson

Immunoresearch, West Grove, Pennsylvania, USA)

Mouse gamma globulins (cat. no: 015-000-003 Jackson

Immunoresearch, West Grove, Pennsylvania, USA)

Peroxidase-conjugated goat anti-mouse IgG (cat. no: 115 035 146, Jackson Immunoresearch, West Grove, Pennsylvania, USA)

Peroxidase-conjugated goat anti-rabbit IgG (cat. no: 111 035 144, Jackson Immunoresearch, West Grove, Pennsylvania, USA)

3.6 GELS

Criterion Tris-HCl Gel, 4–20% polyacrylamide (cat.no: 345-0033, Bio Rad, Hercules, California, USA)

Criterion Tris-HCl Gel, 7.5% polyacrylamide (cat.no:345-0005, Bio Rad, Hercules, California, USA)

Gelfree 8100 Cartridge - 5% tris-acetate (cat.no: 42402, Expedeon, Cambridgeshire, United Kingdom)

(36)

35 Gelfree 8100 Cartridge - 10% tris-acetate (cat.no: 42404, Expedeon, Cambridgeshire,

United Kingdom)

3.7 VARIOUS EQUIPMENT

5mLtube (Sarstedt, Nümbrecht, Germany)

50mL tube (Corning incorporated, New York, USA)

Barseal (Thermo Fisher Scientific, Rockford, Illinois,

USA)

BD HTS Option for BD LSR II (BD Biosciences, San Jose, California, USA)

BD LSR II (BD Biosciences, San Jose, California, USA)

Canon EOS 450D (Canon, Tokyo, Japan)

Criterion Cell (Bio Rad, Hercules, California, USA)

Direct Detect Spectrometer (Merck Millipore, Darmstadt, Germany) Direct Detect Assay-free Cards (Merck Millipore, Darmstadt, Germany)

Centrifuge 5810R, (Eppendorf, Hamburg, Germany)

Gelfree 8100 Fractionation Station (Expedeon, Cambridgeshire, United Kingdom)

Heat block (DRI-block, Techne, Staffordshire, United

Kingdom)

Microtube (Sarstedt, Nümbrecht, Germany)

(37)

36 Microtube centrifuge (Mikro 22R, Hettich zentrifugen, Tuttlingen

Germany)

Microtitre plate 96 well/v-bottom non sterile (Sterilin, Newport, UK)

Mix Mate (Eppendorf, Hamburg, Germany)

Omicron-Laserage Laserprodukte GmbH (Omicron-Laserage, Rodgau, Germany)

PCR tube (Axygen, Corning incorporated, New York,

USA)

Liquid handling robot (Zephyr Compact, Caliper LifeScienses, Hopkinton, Massachusetts, USA)

PowerPac Basic Power Supply (Bio Rad, Hercules, California, USA)

Rotator (Rotator SB3, Stuart, Staffordshire, United

Kingdom)

Sapphire561 DS (Coherent, Santa Clara, California, USA)

3.8 SOFTWARE

Cluster 3.0 ⁸⁷ (Stanford University, Stanford California,

USA)

Excel 2003 (Microsoft, Redmond, Washington, USA)

FACS Diva software (BD Biosciences, San Jose, California, USA) Java TreeView ⁸⁷ (Stanford University, Stanford California,

USA)

(38)

37

4.0 METHODS

4.1 PREPARATION OF WHOLE CELL LYSATES

For the cellular proteins to be accessible for analysis, the cells have to be lysed. In this thesis whole lysates, containing both cytosolic-, nuclear-, organelle - and membrane proteins, were used. This lysate was prepared by suspending the cells in a solution containing NaCl and lauryl maltoside (LM). LM is a strong detergent which dissolves all membranes, and the proteins from organelles, cytosol and membranes will be solubilised. Recently obtained results from our laboratory have shown that a large number of nuclear proteins are released when cells are lysed in buffers containing 140mM NaCl (no yet published).

The cells were cultured in RPMI-medium with 5% foetal bovine serum. When preparing whole cell lysates the cells were first transferred to a 50mL tube and pelleted by centrifugation at 300g for 3 minutes at 4°C. The cells were washed twice in PBS to remove residual protein from the medium.

Approximately 5x10⁷ cells were suspended in 700ul of prepared lysis buffer. The microtube was placed in a 50mL tube prefilled with wet ice and rotated for 30 minutes. After incubation the sample was centrifuged at 24100g at 4°C for 5 minutes. The supernatant was aliquoted to microtubes, and stored at -70°C.

4.2 MEASURING PROTEIN CONCENTRATION

To measure protein concentration, Direct Detect from Millipore was used. This is an infrared- based system that measures the amide bonds in the protein chains and therefore does not rely on the amino acid composition of the proteins, as traditional Bradford assay does ^{88, 89}. Detergents, such as SDS, would interfere with the results obtained with Bradford assays ⁸⁹, and this would be a problem in the experiments described in this thesis.

The sample was applied to the Direct Detect Assay-free Cards. A total volume of 2μL sample was applied to the spots on the card, and the samples where spotted in triplicates. The lysis buffer was used as a blank. The cards were put in the Direct Detect spectrometer, and the proteins concentration was measured.

A multiplexed antibody-based approach for analysis of denatured cellular proteins : development of western-MAP

TABLE OF CONTENTS

ACKNOWLEDGEMENTS

ABSTRACT

SAMMENDRAG

1.0 INTRODUCTION

2.0 BACKGROUND

2.1 HUMAN GENOMICS

2.2 HUMAN PROTEOMICS

2.3 PROTEIN SEPARATION

2.4 ANTIBODY-BASED METHODS

2.5 MASS SPECTROMETRY

2.6 ANTIBODY ARRAY ANALYSIS

3.0 MATERIALS

3.1 CELL LINES

3.3 COMMERCIAL SOLUTIONS

3.4 SOLUTIONS PREPARED IN THE LAB

3.5 ANTIBODIES

3.6 GELS

3.7 VARIOUS EQUIPMENT

3.8 SOFTWARE

4.0 METHODS

4.1 PREPARATION OF WHOLE CELL LYSATES

4.2 MEASURING PROTEIN CONCENTRATION